{
  "id": "1807.04162",
  "version": "v1",
  "published": "2018-07-11T14:39:17.000Z",
  "updated": "2018-07-11T14:39:17.000Z",
  "title": "TherML: Thermodynamics of Machine Learning",
  "authors": [
    "Alexander A. Alemi",
    "Ian Fischer"
  ],
  "comment": "Presented at the ICML 2018 workshop on Theoretical Foundations and Applications of Deep Generative Models",
  "categories": [
    "cs.LG",
    "cond-mat.stat-mech",
    "stat.ML"
  ],
  "abstract": "In this work we offer a framework for reasoning about a wide class of existing objectives in machine learning. We develop a formal correspondence between this work and thermodynamics and discuss its implications.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2018-07-11T14:39:17.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "machine learning",
      "thermodynamics",
      "wide class",
      "formal correspondence"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}